Land Restitution Evaluation Study

SALDRU has been commissioned by the International Initiative for Impact Evaluation (3ie) and the Department of Rural Development and Land Reform (DRDLR) to undertake a comprehensive evaluation of the impacts of the Land Restitution Programme on the economic, social and psychological outcomes of the beneficiaries. To achieve these objectives, SALDRU has launched the Land Restitution Evaluation Study (LRES), which will collect and analyse primary data on beneficiaries before and after they receive their compensations.

The LRES survey will track and interview approximately 6000 land restitution beneficiary households twice over four years. The purpose of the survey is to collect data on a broad range of outcomes, including consumption, food security, production and land use, income, employment, assets, credit, savings, investment, mental health, risk taking, time discounting and cognitive functioning.

Lead Principal Investigator: Malcolm Keswell
Principle Investigator: Michael Carter
Principle Investigator: Mvuselelo Ngcoya

Context and Relevance
Between 1913 and 1994, millions of non-white South Africans were evicted and forcibly relocated both in rural and urban settings in South Africa. This history of land loss was institutionalised by the Natives Land Act of 1913 and other related laws that allocated approximately 87% of land to whites and reserved the remaining land for blacks. Through this process of evictions and forced relocations, much of the fertile and productive land came under the control of White South Africans, leaving the Black population with marginal and relatively unproductive land (Walker, 2014). This was not restricted to rural agricultural regions; in urban areas evictions sought to homogenise neighbourhoods and often moved blacks to the periphery of the cities. It is estimated that at the height of the Apartheid years (1960-1983), over 3 million (out of an estimated 20-24 million) non-white South Africans were removed from their lands by the state (Platzy & Walker, 1985). Furthermore, when including the removals due to other racially charged land consolidation policies, the number of dispossessed is estimated to be over 7 million. In 1994 the Restitution of Land Rights Act was passed. This Act entitled individuals and their descendants who were dispossessed of their land by racially motivated legal acts to file claims for restitution. The key output of this Land Restitution programmme is a sizeable once-off cash or land transfer to individuals. These transfers are large enough to have the potential to lift households out of poverty into a virtuous cycle leading to higher productivity and improved long-term wellbeing. However, there is no quantitative evidence to date on the impacts of this programme for poverty alleviation or even whether the Act’s goal of “restorative justice” was achieved. This
evaluation will, in the first instance, provide the much needed answers to these basic first-order questions of interest to both the implementing agency as well as to policy makers in general.

Evaluation Questions of Interest
Evidence in other countries on the effects of modest asset transfers is generally positive, but there are very few studies that attempt to identify the effects of large cash or asset transfers such as this one. Two notable exceptions { Keswell & Carter (2014) and Haushofer and Shapiro (2016) } show large and lasting gains on a number of different outcomes. This study will considerably broaden our understanding of how and under what conditions large cash or asset transfers work and whether impacts are mediated by psychological and behavioural pathways. The key questions we will seek to address are:

  • What are the impacts of the programme on consumption, production, investment and savings?
  • Do land transfers lead to increases in the returns to labour, entrepreneurial and farming skills?
  • Does the programme alter the risk preferences and time discounting behaviour of the beneficiaries with consequences for the uptake of credit and propensity to save and invest?
  • Can a large once-off cash transfer have an impact on executive functioning (the set of .cognitive processes that are central to cognition, inhibition, memory, problem solving and planning (Haushofer and Shapiro, 2016; Haushofer and Fehr, 2016; Manni, Mullainathan, Shar and Zhao, 2013) – and is this an important pathway out of poverty?
  • What is the impact of land transfers on social cohesion? For the purposes of this study, we dene social cohesion over both individual as well as group outcomes. On an individual level, we will look at impacts on trust, altruism, aspirations, positive affect, happiness, and political attitudes. At the group level, we will look at impacts on collective action and civic engagement broadly.
  • Are the impacts on consumption and production sustained?

Empirical Data and Methods
There are two sub-populations of beneficiaries: claimants that elect to receive cash compensation and claimants that elect to have their rights to the land restored (or rights to alternative lands). Our evaluation draws a sample from the population of claims still in the pipeline to study the impacts of the cash or land transfers. We use a pipeline randomisation identification strategy where for each output we randomise the timing of the transfer into “early” and “late” treatment arms, where the claimants that receive their transfers later would serve as the control group to claimants that receive their transfers earlier. Our data collection iterates between quantitative and qualitative research methods starting with pre-baseline qualitative fieldwork on a subset of land reform claimants who will not form part of the study sample in order to sharpen the various themes and hypotheses to be investigated. We also conduct follow-up surveys of all study participants, including qualitative follow-up interviews with a subset of study participants. Finally, we employ a mix of survey and task-based methods to measure executive functioning, risk, time discounting, and other dimensions of preferences. The sample sizes that are drawn to study the impacts of the land and cash transfer outputs of the programme are 2800 and 3250 households respectively. These samples are randomised into the early and late treatment arms in equal proportion. Our follow up surveys takes account of households that move or split. These sample sizes are powered to detect policy relevant effect sizes on consumption.